Medicare-certified home care agencies are under increasing pressure to improve both resource efficiency and quality of care for patients who are insured by the Medicare program. Home care agency clinicians and managers, predominantly nurses, are responding by collecting and reviewing standardized data to monitor resource use and patient outcomes. Yet little scientific knowledge has been produced to explain how outcomes in Medicare home care patients are associated with the amount of clinical staff resources allocated for these patients during home care episodes. Therefore, the long-term objective of this study is to improve our understanding of the complex interplay between resource use and health-related outcomes in Medicare home care patients with diverse characteristics who are served by a diverse group of Medicare-certified home care agencies. The primary aim of this study is to determine how home care patient outcomes are associated with specific home care resource use measures, controlling for market, home care agency, and patient variables. The secondary aim is to determine the relative importance of market factors, agency characteristics, patient characteristics, and home care resource use in influencing home care patient outcomes. Hypothesis 1: Patients who end an episode of care due to hospitalization will use the same amount of home care resources as patients who end an episode of care and remain at home, but they will show greater service intensity, controlling for market, agency, and patient variables. Hypothesis 2: Patients who decline in functional status during an episode of home care will use the same amount of home care resources as patients who improve in functional status, controlling for market, agency, and patient variables. Hypothesis 3: Patients whose symptoms worsen during an episode of home care will use the same amount of home care resources as patients whose symptoms improve, controlling for market, agency, and patient variables. Hypothesis 4: The observed relationships (direct and indirect effects) between market factors, agency characteristics, home care resource use, and home care patient outcomes will be adequately represented by a structural equation model. Using a prospective cohort study design, we will enroll 1,500 patients from 30 randomly-selected Ohio home care agencies. Patient data will be collected by trained home care staff, based on assessments at the first and final visits of episodes. Information about resource use at every home visit will also be recorded. Data analyses will determine unadjusted and adjusted relationships between measures of resource use and patient outcomes. A Project Advisory Committee will help translate findings into strategies for planning home care resource allocation based on patient outcomes.